f-string-in-exception (EM102)

Derived from the flake8-errmsg linter.

Fix is sometimes available.

What it does

Checks for the use of f-strings in exception constructors.

Why is this bad?

Python includes the raise in the default traceback (and formatters like Rich and IPython do too).

By using an f-string, the error message will be duplicated in the traceback, which can make the traceback less readable.

Example

Given:

sub = "Some value"
raise RuntimeError(f"{sub!r} is incorrect")

Python will produce a traceback like:

Traceback (most recent call last):
  File "tmp.py", line 2, in <module>
    raise RuntimeError(f"{sub!r} is incorrect")
RuntimeError: 'Some value' is incorrect

Instead, assign the string to a variable:

sub = "Some value"
msg = f"{sub!r} is incorrect"
raise RuntimeError(msg)

Which will produce a traceback like:

Traceback (most recent call last):
  File "tmp.py", line 3, in <module>
    raise RuntimeError(msg)
RuntimeError: 'Some value' is incorrect